Predicting soil strength with remote sensing data

• Main Author: Wende, Jon T.
• Other Authors: Olsen, R.C.
• Published: Monterey, California. Naval Postgraduate School 2012
• Online Access: http://hdl.handle.net/10945/5174

Approved for public release; distribution is unlimited === Predicting soil strength from hyperspecral imagery enables amphibious planners to determine trafficability in the littorals. Trafficability maps can then be generated and used during the intelligence preparation of the battlespace allowing amphibious planners to select a suitable landing zone. In February and March 2010, the Naval Research Laboratory sponsored a multi-sensor remote sensing and field calibration and field validation campaign (CNMI'10). The team traveled to the islands of Pagan, Tinian, and Guam located in the Marianas archipelago. Airborne hyperspectral imagery along with ground truth data was collected from shallow water lagoons, beachfronts, vegetation, and anomalies such as World War II relics. In this thesis, beachfront hyperspectral data obtained on site was used as a reference library for evaluation against airborne hyperspectral data and ground truth data in order to determine soil strength for creating trafficability maps. Evaluation of the airborne hyperspectral images was accomplished by comparing the reference library spectra to the airborne images. The spectral angle between the reference library and airborne images was calculated producing the trafficability maps amphibious planners can use during the intelligence preparation of the battlespace.